The forward and feedback information may be represented by log-likelihood ratios.
Techniques for efficiently and accurately computing log-likelihood ratio (LLRs) for code bits are described.
In the receiving device, LLR (Log-Likelihood Ratio) values are stored into an LLR buffer.
The invention comprises steps (340) consisting of determining a simplified expression specific to the logarithm of the likelihood ratio.
A log-likelihood ratio calculation module (18) is configured to compute bit-level LLRs from the coefficients and the received signal.
One version of the algorithm monotonically decreases the negative log-likelihood cost function each iteration.
A parity bit is estimated from the logarithmic likelyhood ratio of the parity bit.
The signal quality used is an intrinsic signal-to-noise ratio derived from the log-likelihood-ratio of the soft outputs of the decoded window.
The invention also provides a refinement mechanism that is operative to significantly improve the log likelihood (LLR) of the list of candidates.
Lower information bits are estimated from the logarithmic likelyhood ratio of the lower information bits.
Date is received without NULLs and inputted into a log-likelihood ratio (LLR) combining block.
A decoding unit (413) decodes the received signal and outputs the post log likelihood ratio (LLR) of the encoded bits.
The method selects a mismatched energy ratio (k2) between k12 and k22, and generating bit LLR values for two-layered QPSK turbo decoding, using the mismatched k2 energy ratio.